This repository contains an exploratory data analysis (EDA) of a comprehensive dataset on ultra-marathon running events. The dataset spans over two centuries and provides detailed information on various ultra-marathon races, athletes, and their performances.
The dataset used in this analysis is sourced from Kaggle: The Big Dataset of Ultra-Marathon Running. It includes information on race events, distances, athlete demographics, and performance metrics.
The Jupyter notebook EDA_marathon_running.ipynb
covers the following steps:
-
Data Acquisition
- Downloading the dataset using the Kaggle API.
- Extracting the data from the downloaded zip file.
-
Data Preprocessing
- Loading the dataset into a pandas DataFrame.
- Initial exploration of the dataset (shape, data types, missing values).
- Filtering the dataset for specific criteria (e.g., events in 2020, races held in the USA).
- Cleaning and transforming data columns for analysis.
-
Exploratory Data Analysis (EDA)
- Visualizing the distribution of race distances and athlete genders.
- Analyzing the relationship between athlete age and performance.
- Comparing average speeds across different event distances and genders.
Several visualizations are generated to understand the data better:
- Count plots for event distances and athlete genders.
- Violin plots for athlete speeds across event distances.
- Line plots showing the relationship between athlete age and average speed.
To run the notebook, you need the following Python libraries:
- pandas
- seaborn
- matplotlib
- kaggle
This analysis provides insights into ultra-marathon running events, highlighting trends and patterns in athlete performances based on gender, age, and event distances. Further analysis can be conducted to explore additional aspects of the dataset.